Bittensor (TAO) Returns by Hour, Weekday & Session
When has Bittensor tended to rise or fall? Average returns by hour of day (UTC), weekday and trading session, measured on Binance spot TAOUSDT hourly closes from 11 Apr 2024 to 5 Oct 2026. Seasonality in the past is not a forecast: most of these differences are small next to an hour's normal move.
Over 21,779 hourly returns since 11 Apr 2024, Bittensor's strongest UTC hour on average was 21:00 (+0.067%, up 51.0% of the time) and its weakest 10:00 (-0.105%); the best hour's edge is within the noise once you allow for picking the best of 24 hours (t = 1.7). By UTC day, Saturday averaged the most (+0.45%) and Tuesday the least (-0.58%), over 907 days. Weekends are quieter: the average weekend day ranged 6.37% against 7.73% on weekdays (17% less). Of the trading sessions, the Asia session added the most in total (+16.8% cumulative over 647 weekdays).
Bittensor average return by hour of day (UTC)
11 Apr 2024 – 5 Oct 2026 · 21779 hourly returns| Hour (UTC) | Mean | Median | % up | Avg |move| | t-stat | Hours |
|---|---|---|---|---|---|---|
| 00:00 | +0.024% | +0.027% | 50.4% | 0.875% | +0.6 | 907 |
| 01:00 | -0.064% | -0.060% | 46.3% | 0.883% | -1.5 | 907 |
| 02:00 | +0.061% | +0.044% | 52.0% | 0.806% | +1.6 | 907 |
| 03:00 | +0.037% | -0.028% | 48.4% | 0.738% | +1.1 | 907 |
| 04:00 | +0.036% | -0.023% | 47.6% | 0.780% | +1.0 | 907 |
| 05:00 | +0.033% | +0.000% | 49.8% | 0.753% | +0.9 | 907 |
| 06:00 | -0.027% | -0.024% | 47.0% | 0.815% | -0.7 | 907 |
| 07:00 | +0.007% | +0.000% | 48.3% | 0.797% | +0.2 | 907 |
| 08:00 | +0.003% | +0.000% | 49.2% | 0.838% | +0.1 | 907 |
| 09:00 | +0.015% | -0.028% | 47.9% | 0.797% | +0.4 | 907 |
| 10:00 | -0.105% | -0.111% | 43.4% | 0.744% | -3.1 | 907 |
| 11:00 | +0.023% | +0.000% | 49.1% | 0.742% | +0.6 | 907 |
| 12:00 | -0.039% | -0.080% | 45.4% | 0.872% | -0.9 | 907 |
| 13:00 | -0.020% | -0.011% | 49.1% | 1.014% | -0.4 | 908 |
| 14:00 | +0.045% | +0.000% | 49.2% | 1.088% | +0.9 | 908 |
| 15:00 | -0.027% | -0.039% | 47.6% | 1.029% | -0.6 | 908 |
| 16:00 | -0.031% | -0.034% | 48.5% | 0.913% | -0.7 | 908 |
| 17:00 | +0.005% | +0.000% | 49.1% | 0.938% | +0.1 | 908 |
| 18:00 | -0.069% | -0.081% | 46.5% | 0.885% | -1.7 | 908 |
| 19:00 | -0.005% | -0.033% | 47.2% | 0.852% | -0.1 | 908 |
| 20:00 | -0.005% | +0.000% | 49.8% | 0.823% | -0.1 | 908 |
| 21:00 | +0.067% | +0.038% | 51.0% | 0.802% | +1.7 | 908 |
| 22:00 | +0.005% | -0.010% | 48.3% | 0.808% | +0.1 | 908 |
| 23:00 | -0.057% | -0.029% | 47.1% | 0.692% | -1.8 | 908 |
Bittensor average return by day of the week
UTC-day closes · 907 days| Weekday | Mean | Median | % up | Avg |move| | t-stat | Days |
|---|---|---|---|---|---|---|
| Monday | +0.319% | -0.432% | 47.7% | 4.475% | +0.6 | 130 |
| Tuesday | -0.578% | -0.808% | 45.7% | 4.262% | -1.2 | 129 |
| Wednesday | -0.271% | -0.814% | 42.6% | 3.857% | -0.6 | 129 |
| Thursday | -0.493% | -1.114% | 46.5% | 4.391% | -1.0 | 129 |
| Friday | -0.108% | -0.428% | 45.4% | 4.465% | -0.2 | 130 |
| Saturday | +0.449% | -0.180% | 46.9% | 3.279% | +1.1 | 130 |
| Sunday | +0.129% | -0.274% | 45.4% | 3.762% | +0.3 | 130 |
Bittensor return by trading session
Weekdays · sessions overlap| Session | Hours | Mean | Median | % up | Avg |move| | Cumulative | Sessions |
|---|---|---|---|---|---|---|---|
| Asia | 00:00–08:00 UTC | +0.024% | +0.031% | 50.2% | 2.34% | +16.8% | 647 |
| Europe | 07:00–16:00 London | -0.170% | -0.269% | 46.1% | 2.68% | -66.8% | 647 |
| US | 09:00–16:00 New York (NYSE opens 09:30) | -0.198% | -0.321% | 45.5% | 2.56% | -72.3% | 648 |
The sessions overlap (Asia and Europe share an hour or two, Europe and the US two to three), so their returns don't add up to the day's. Cumulative is every session's return compounded over the window.
Bittensor weekend vs weekday volatility
UTC days| Days | Avg |daily return| | Avg daily range | Avg |hourly return| | Days |
|---|---|---|---|---|
| Weekdays (Mon–Fri) | 4.29% | 7.73% | 0.893% | 647 |
| Weekend (Sat–Sun) | 3.52% | 6.37% | 0.726% | 260 |
Range is the highest over the lowest hourly close of a UTC day (days with at least 20 stored hours).
Bittensor hour × weekday heatmap
Mean hourly return in basis points (1 bp = 0.01%) · UTC| Day | 00 | 01 | 02 | 03 | 04 | 05 | 06 | 07 | 08 | 09 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mon | +10.4 | -25.8 | +7.3 | +4.7 | +7.0 | +10.7 | -19.5 | +12.3 | -0.9 | +3.5 | -20.1 | +6.5 | -2.5 | -5.4 | +18.6 | -10.6 | +10.0 | +1.1 | -7.7 | +25.8 | +10.2 | -2.6 | +7.4 | -8.4 |
| Tue | -3.1 | -20.9 | +0.7 | +2.8 | -12.3 | +9.8 | -11.3 | -8.4 | +3.3 | +0.0 | +0.3 | +6.0 | -20.0 | -11.9 | +6.6 | +3.4 | -6.4 | -9.2 | +4.4 | -2.7 | -3.3 | +11.2 | +1.7 | +1.5 |
| Wed | -3.4 | -1.5 | -2.6 | -5.7 | +12.0 | -1.1 | +14.2 | +8.8 | -0.1 | +0.8 | -4.2 | -6.0 | -6.7 | -4.2 | +3.8 | -3.7 | -15.1 | +7.3 | -6.7 | -5.8 | +10.4 | +9.9 | -18.4 | -8.9 |
| Thu | +23.0 | -4.7 | +15.5 | -10.5 | +5.3 | -0.4 | +0.3 | -7.1 | +2.1 | -12.6 | -5.3 | +0.1 | -4.3 | -8.3 | -14.4 | -11.2 | -2.1 | -2.5 | -15.2 | -18.4 | +6.6 | +21.9 | -3.3 | -11.4 |
| Fri | -11.5 | +17.7 | -4.8 | +10.5 | +8.7 | +0.7 | -9.2 | -6.5 | +9.4 | +15.5 | -11.6 | +0.1 | +13.3 | +12.4 | -11.9 | -5.1 | +2.4 | -4.1 | -26.9 | +9.2 | -9.1 | -8.1 | -5.0 | +3.2 |
| Sat | -0.5 | -3.3 | +27.7 | +13.0 | +6.2 | -0.9 | +2.0 | -5.0 | -8.8 | +5.3 | -12.5 | -2.1 | -0.9 | -6.2 | +13.4 | -1.4 | -9.6 | +2.9 | +6.1 | -14.2 | -2.3 | +26.0 | +9.1 | +0.9 |
| Sun | +2.3 | -6.6 | -0.7 | +10.8 | -1.9 | +4.4 | +4.7 | +10.8 | -2.9 | -2.0 | -19.9 | +11.5 | -6.3 | +9.7 | +15.4 | +9.4 | -1.2 | +7.9 | -2.2 | +2.9 | -16.3 | -11.7 | +11.6 | -16.8 |
How to read it. Mean and median are average hourly (or daily, or session) returns; % up is the share that closed higher; Avg |move| is the average size of the move either way. The best of 24 hours will look good by luck alone, so an hour needs a |t-stat| of about 3.1 or more (about 2.7 for the best of 7 weekdays) before its average is distinguishable from zero; |t| of 2 is not enough here. Returns are only counted where both closes are stored: 0 missing hours in this window were skipped, not bridged. See also Bittensor monthly returns.
Source: Binance spot TAOUSDT hourly candles (public API). Past seasonality is not a forecast. Not financial advice.
Bittensor seasonality FAQ
What is the best time of day to buy Bittensor?
The tables show which UTC hours and weekdays have had the highest and lowest average return in the data, with the share of hours that closed up and a t-statistic. Most hourly differences are tiny (hundredths of a percent) and within the noise. Because the best of 24 hours is picked after the fact, a single hour needs |t| of about 3.1 or more (a Bonferroni correction for 24 comparisons; about 2.7 for the best of 7 weekdays) before it counts as more than luck. Past seasonality is not a forecast.
How are the returns calculated?
From Binance spot USDT hourly candles: each hour's return is the log change from the previous hour's close to its own close, and each weekday's return from one UTC-day close (the 23:00 candle) to the next. A return is only counted when both closes are stored; Binance skips the odd hour for maintenance, and those gaps are left out rather than bridged. Percentages are converted back to simple returns for display.
What are the trading sessions?
Asia is 00:00–08:00 UTC; Europe 07:00–16:00 London time (06:00–15:00 UTC in British Summer Time); the US is the NYSE session, 09:30–16:00 New York time, measured 09:00–16:00 because the candles are hourly (13:00–20:00 UTC in US daylight time, 14:00–21:00 otherwise). The sessions overlap, so their returns do not add up to the day's. Weekdays only; exchange holidays are not taken out.
Is crypto more volatile on weekends?
The weekend table compares the average absolute daily return, the average intraday range (highest over lowest hourly close in a UTC day) and the average absolute hourly return on Saturdays and Sundays with Monday to Friday.